[Question]: Cutoff Guidance on "threshold on row means of normalized counts"

[tamuanand]

Hi @federicomarini

Thanks a lot for the great package.

In the "counts overview tab", I wanted to seek guidance on the recommended cutoff to use when filtering based on "threshold on the row means of the normalized counts".

• when it comes to "threshold on the row sums of the counts", there is some kind of guidance here - https://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html#pre-filtering which states - Here we perform pre-filtering to keep only rows that have a count of at least 10 for a minimal number of samples. The count of 10 is a reasonable choice for bulk RNA-seq

I am trying to find any literature/guidance if I use "threshold on row means of normalized counts".

Thanks in advance.

[federicomarini]

Hi @tamuanand , I think that is pretty much an arbitrary choice that

• lets you focus on the genes meaningful to you, avoiding focusing on low expressed genes
• lets you spend efficiently your "FDR budget" on the features that have better chance to be detected as DE

I tend to (re)read myself the vignette of DESeq2 in these cases.
HTH
Federico